Ignoring contact list corruption and overwriting isn’t really a solution to the problem
Discussion
problem with nip65 is that its not just a relay list spec, it prescribes a specific way to pull data from these relays which damus isn’t really setup to do (yet).
Are you talking about the read and write optional tags on the specified relays?
Yes to me that NIP is more a signal that the client uses the gossip model than anything else. Doesn’t add anything otherwise.
I’d love to understand how you’d prefer users set and publish their preferred relays. The kind 3 contact list seems a little brittle to me
how so
I vehemently reject the idea that nostr clients have to keep implementing the “next great idea” so that it keeps working. Bcash mentality.
I agree, I don’t want that culture either. Just looking at the nip for the kind 3 event it doesn’t look like it was intended to be where people published their preferred relays.
I’d be happy to standardise on kind 3 if we updated the NIP for that expressed purpose. But in talking to some folks they thought NIP 65 would be the way to solve the preferred relay problem going forward.
I just wanna go with what people prefer I am agnostic. Just pulling for one standard and I haven’t heard consensus yet.
This conviction is the best thing about you.